Barth Syndrome

A number sign (#) is used with this entry because Barth syndrome, also known as 3-methylglutaconic aciduria type II (MGCA2), is caused by mutation in the tafazzin gene (TAZ; 300394) on chromosome Xq28.

Description

Barth syndrome (BTHS) is an X-linked disease conventionally characterized by dilated cardiomyopathy (CMD) with endocardial fibroelastosis (EFE), a predominantly proximal skeletal myopathy, growth retardation, neutropenia, and organic aciduria, particularly excess of 3-methylglutaconic acid. Features of the disease that are less well known include hypertrophic cardiomyopathy, isolated left ventricular noncompaction (LVNC), ventricular arrhythmia, motor delay, poor appetite, fatigue and exercise intolerance, hypoglycemia, lactic acidosis, hyperammonemia, and dramatic late catch-up growth after growth delay throughout childhood (summary by Steward et al., 2010).

For a phenotypic description and a discussion of genetic heterogeneity of 3-methylglutaconic aciduria, see MGCA type I (250950).

Clinical Features

Barth et al. (1981, 1983) described a large Dutch pedigree showing X-linked inheritance of a disorder characterized by dilated cardiomyopathy, neutropenia, skeletal myopathy, and abnormal mitochondria. By electron microscopy, the mitochondria showed concentric, tightly packed cristae and occasional inclusion bodies.

Hodgson et al. (1987) thought that the same disorder was present in the family they reported in which many males in at least 3 generations and 7 sibships connected through females died between ages 3 days and 31 months of sepsis due to agranulocytosis or of cardiac failure. Weakness of skeletal muscles with sparing of the extraocular and bulbar muscles was noted. Granulocytopenia was found as early as cord blood samples. Differentiation in the bone marrow was arrested at the myelocyte stage. None of the boys had a gross structural cardiac abnormality. Endocardial fibroelastosis was documented in 2, and in 1 of these, electron microscopy demonstrated abnormality of mitochondria.

Hodgson et al. (1987) also suggested that the family reported by Neustein et al. (1979) had the same disorder. Neustein et al. (1979) demonstrated abnormal mitochondria on electron microscopic examination of a transvascular endomyocardial biopsy from an infant with cardiomyopathy and chronic congestive heart failure. At autopsy, similar abnormal mitochondria were seen in skeletal muscle, liver, and kidneys. In 3 other males in 2 sibships related as first cousins or first cousins once removed, autopsy showed endocardial fibroelastosis and, by electron microscopy, abnormal mitochondria. A heterozygote showed no abnormality on skeletal muscle biopsy. No mention of neutropenia in the affected males was made.

Ino et al. (1988) reported cases of dilated cardiomyopathy, short stature, and abnormal carnitine metabolism.

Fixler et al. (1970) described 4 males in 3 sibships, related through females, with the contracted form of endocardial fibroelastosis, which is frequently associated with malformations of the heart. The affected males died of heart failure in the first years of life. Lindenbaum et al. (1973) described a British kindred in which there were 2 males over 2 generations with endocardial fibroelastosis. The propositus and a male first cousin of his mother died in infancy of 'heart trouble.' Autopsies on both confirmed the primary dilated type of endocardial fibroelastosis. One had no other birth defects; the other had a hypoplastic left kidney. Several other males of this kindred died before the age of 2 years. This pattern of inheritance, along with the findings of Fixler et al. (1970), suggested X-linked transmission. Westwood et al. (1975) described a family with a pedigree consistent with X-linked recessive inheritance in 3 males in successive generations.

Kelley et al. (1989, 1991) elaborated on the clinical picture of this disorder on the basis of 7 affected boys form 5 unrelated families with dilated cardiomyopathy, growth retardation, neutropenia, and persistently elevated urinary levels of 3-methylglutaconate, 3-methylglutarate, and 2-ethylhydracrylate. The clinical course of the disorder was characterized by severe or lethal cardiac disease and recurrent infections during infancy and early childhood but relative improvement in later childhood. The initial presentation of the syndrome varied from congenital dilated cardiomyopathy to infantile congestive heart failure to isolated neutropenia without clinical evidence of heart disease. The excretion of 3-methylglutaconate and 3-methylglutarate appeared to be independent of the metabolism of leucine, the presumed precursor of these organic acids. Chitayat et al. (1992) referred to this form of 3-methylglutaconic aciduria as type II.

Orstavik et al. (1993) reported 3 families with possible X-linked congestive cardiomyopathy associated with specific abnormalities of the mitochondria. The heart disorder presented as endocardial fibroelastosis with neonatal death in 2 brothers in 1 family and as heart failure and death in infancy in 2 brothers in the other 2 families. In 1 family, a maternal uncle may also have been affected. Pyoderma and neutropenia were reported in 1 of the boys. Electron microscopy of heart muscle showed increased numbers of mitochondria and abnormal mitochondrial crystal condensations and paracrystalline inclusions in all sibships.

Ades et al. (1993) studied a large Australian family with no known Dutch forebears in which affected males over 3 generations had dilated cardiomyopathy, short stature, and neutropenia. Age at diagnosis ranged from 6 weeks to 10 years, with a maximum recorded survival age of 10 years and 3 months. Clinical details were available for 6 boys, 4 deceased and 2 living. Cardiomyopathy and progressive growth failure with decline of both length and weight velocities over time were the most consistent clinical markers of disease. Some patients displayed endocardial fibroelastosis. Neutropenia was congenital and persistent in 1 boy, recurrent in 2, and documented once in another. Skeletal myopathy was present in 3 boys and was heralded by delay in gross motor development or an abnormal gait. One boy had clinical peripheral neuropathy and complex neuroophthalmologic signs, suggesting involvement of the lower midbrain and possibly the cerebellum. Ades et al. (1993) noted that ophthalmoplegia is a recognized finding in mitochondrial myopathies, but had not previously been reported in patients with Barth syndrome. Additional findings included talipes equinovarus in 2 boys, 1 of whom also had minor facial anomalies and congenital pectus excavatum.

Christodoulou et al. (1994) described 6 cases of Barth syndrome from 4 families, including 5 patients who were still alive at ages 11 months, 2 years, 5.9 years, 6.5 years, and 13 years. The authors noted that neuromuscular and cardiovascular symptoms and severity of infections tended to improve with age, whereas short stature persisted. In addition, they observed myopathic facies and a nasal quality to speech in their cases. Urinary organic acid abnormalities and plasma carnitine deficiency were inconsistent findings.

Gedeon et al. (1995) reported a large Australian family in which male infants died from congenital dilated cardiomyopathy. There was a strong family history of unexplained death in infant males over at least 4 generations in a pattern consistent with X-linked recessive inheritance. Death always occurred in early infancy, without development of the characteristic features associated with Barth syndrome, such as skeletal myopathy, short stature, and neutropenia. Two of the patients also had talipes equinovarus. Affected members of this family were originally thought to have a form of dilated cardiomyopathy, which was designated CMD3A.

Bleyl et al. (1997) described the clinical and pathologic findings of a 4-generation Utah family in which 6 males were affected with severe X-linked cardiomyopathy. Consistent findings included neonatal onset of ventricular dysfunction frequently complicated by arrhythmias and cardiac failure during the first year. Growth retardation was seen in 4 of the patients, neutropenia was seen in 2, and 1 patient had muscle weakness. Electrocardiographic findings were diagnostic of isolated noncompaction of the left ventricular myocardium (LVNC) (Chin et al., 1990). Fetal echocardiograms obtained between 24 to 30 weeks' gestation in 3 of the affected males showed a dilated left ventricle in 1, but were not otherwise diagnostic of LVNC in any of the patients. Four of the affected individuals died during infancy, 1 was in cardiac failure at age 8 months, and 1 was alive following cardiac transplant at age 9 months. The hearts from the infants who died or underwent transplantation showed dilation and hypertrophy, with coarse, deep ventricular trabeculations within the left ventricle, and prominent endocardial fibroelastosis, characteristic of LVNC. Histologically, the myocardium showed loosely arranged fascicles of myocytes, especially in the subepicardial regions and more prominent in the left ventricle. Markedly elongated mitochondria were present in some ventricular myocytes. With cardiac transplantation, a patient had survived to the age of 7 years at the time of report; with aggressive medical management, another patient was alive at age 14 months.

Marziliano et al. (2007) reported a 12-year-old boy with Barth syndrome. The boy had left ventricular noncompaction and dilated cardiomyopathy, which was detected at 3 months, skeletal myopathy, recurrent oral aphthous ulcers, and cyclic neutropenia. Left ventricular function progressively improved from age 5 years and became subclinical and normal; he presented at age 11 with recurrent ulcers and signs of myopathy, including muscle weakness and atrophy. Molecular analysis identified a mutation in the TAZ gene (300394.0012) inherited from his unaffected mother. He was also heterozygous for a mutation in the LDB3 gene (605906), which is associated with left ventricular noncompaction. The patient's father and brother also carried the LDB3 mutation and had evidence of left ventricular trabeculation on imaging without dysfunction. The significance of the LDB3 mutation was unclear.

Hastings et al. (2009) studied 12 patients from 10 families with mutation-proven Barth syndrome (see, e.g., 300394.0006) and observed similarity in the facial features of the boys. The characteristic facies was most evident in infancy and included a tall and broad forehead, round face with prominent chin and full cheeks, large ears, and deep-set eyes. The features became less evident during puberty and into adulthood, with loss of the prominence of the cheeks. The most striking feature was the development of gynoid stature and fat distribution during the late pubertal period of 'catch-up' growth.

Steward et al. (2010) reported that 6 of 19 UK families with genetically and biochemically proven Barth syndrome (see, e.g., 300394.0006) had male fetal loss and stillbirths in addition to severe neonatal illness or death. In these families, there were multiple miscarriages of male fetuses, 9 males were stillborn, and 14 males died as neonates or infants, but there were no miscarriages, stillbirths, or childhood deaths of females. BTHS was definitively proven in 5 males with fetal onset of CMD with or without hydrops, endocardial fibroelastosis, and/or left ventricular noncompaction. Steward et al. (2010) suggested that Barth syndrome is an underrecognized cause of male fetal demise.

Thompson et al. (2016) conducted a multidisciplinary investigation involving 42 patients with BTHS, including echocardiograms, muscle strength testing, functional exercise capacity testing, physical activity assessments, cardiolipin analysis, 3-methylglutaconic acid analysis, and review of genotype data. Echocardiography revealed considerable variability in cardiac features. By contrast, almost all patients had significantly reduced functional exercise capacity. Multivariate analysis revealed significant relationships between cardiolipin ratio and left ventricular mass and between cardiolipin ratio and functional exercise capacity.

Female Carriers

Female carriers of the BTHS gene appear to be healthy. This could be due to a selection against cells that have the mutant allele on the active X chromosome. Orstavik et al. (1998) therefore analyzed X-chromosome inactivation in 16 obligate carriers of BTHS from 6 families, using PCR of a polymorphic CAG repeat in the first exon of the androgen receptor gene (AR; 313700). An extremely skewed X-inactivation pattern (equal to or more than 95:5), not found in 148 female controls, was demonstrated in 6 carriers. The skewed pattern in 2 carriers from 1 family was confirmed in DNA from cultured fibroblasts. Five carriers from 2 families had a skewed pattern, between 80:20 and less than 95:5, a pattern that was found in only 11 of 148 female controls. Of the 11 carriers with a skewed pattern, the parental origin of the inactive X chromosome was maternal in all 7 cases for which this could be determined. In 2 families, carriers with an extremely skewed pattern and carriers with a random pattern were found. The skewed X inactivation in 11 of 16 carriers is probably the result of a selection against cells with the mutated gene on the active X chromosome. Since BTHS also shows great clinical variation within families, additional factors are likely to influence the expression of the phenotype. Such factors may also influence the selection mechanism in carriers.

Barth (2005) stated that no obligate or genetically proven female carriers had been reported with symptoms of the disease, and the survival of carriers did not differ from the general population.

Reviews

Barth et al. (2004) updated information on Barth syndrome. Following the prediction that the TAZ gene encodes one or more acyltransferases (Neuwald, 1997), lipid studies in patients with Barth syndrome showed a deficiency of cardiolipin, especially its tetralinoleoyl form (L4-CL) (Vreken et al., 2000). Deficiency of L4-CL was subsequently demonstrated in a variety of tissues from patients with Barth syndrome (Schlame et al., 2002), with determination in platelets or cultured skin fibroblasts being the most specific biochemical test. Barth syndrome was the first identified inborn error of metabolism that directly affects cardiolipin, a component of the inner mitochondrial membrane necessary for proper functioning of the electron transport chain. Barth et al. (2004) found that some patients with Barth syndrome have deficient docosahexaenoic acid and arachidonic acid. They pointed out that the initial impression of a uniformly lethal infantile disorder had to be modified. Age distribution in 54 living patients ranged from neonate to 49 years and peaked around puberty. Mortality was highest in the first 4 years. An update on a family with affected members in 3 successive generations and by inference in 2 earlier generations reported by Barth et al. (1983) was provided.

Barth (2005) traced the medical history of X-linked cardioskeletal myopathy and neutropenia (Barth syndrome) to studies in the 1970s that suggested an X-linked mode of inheritance for some families with so-called endocardial fibroelastosis, a term for the pearly-white fibrotic endocardium seen at autopsy in affected individuals; this descriptive term fell into disuse when the emphasis shifted to the study of cardiac dynamics with the advent of echocardiography, with the focus on dilated cardiomyopathy. BTHS commonly presents in infancy with 1 of the following symptoms: failure to thrive, primarily due to dilated cardiomyopathy; delayed motor milestones, with proximal muscle weakness; or bacterial and/or fungal infections due to neutropenia. Barth (2005) noted that some patients reach adult age; however, there is remarkable intrafamilial variability. Cardiomyopathy and neutropenia are the main causes of high mortality, predominantly in the first 5 years of life. Proximal weakness appears to be present from birth; mild facial weakness can be observed, but there are no difficulties with swallowing, eye movements, or ventilation. There is no progression of muscle weakness and no loss of ambulation. A mild learning disability may form part of the disorder. Increased excretion of 3-methylglutaconic acid is the most characteristic biochemical marker of disease, although it is not invariably present. The neutrophil count can vary between normal and zero. Although no longer needed for diagnosis, histochemical analysis of muscle biopsy most commonly shows an increase of sarcoplasmic fat droplets on oil-red-O staining, with minimal changes to mitochondria seen on electron microscopy; heart muscle mitochondria in BTHS exhibit gross changes in shape, size, and alignment of cristae.

Diagnosis

Cantlay et al. (1999) identified 5 unrelated families within a 7-year period in 1 hospital in the area of Bristol, England, with BTHS. Mutations in the G4.5 gene were found in all of the cases (see, e.g., 300394.0006). The authors questioned whether BTHS is underdiagnosed and suggested that all male infants or young children presenting with idiopathic dilated cardiomyopathy be carefully investigated for BTHS. They noted that associated neutropenia is variable, and urinary 3-methylglutaconic acid levels fluctuate. They advocated mutation analysis, if available.

Valianpour et al. (2002) used high performance liquid chromatography-electrospray mass spectrometry to quantify total cardiolipin and molecular subclasses in fibroblasts from 5 patients with Barth syndrome and compared the values to those in a healthy control group and a group with other diseases. Patients with Barth syndrome had decreased total cardiolipins and cardiolipin subclasses, especially tetralinoleoyl-cardiolipin. They suggested use of this biochemical test for diagnosis, followed by mutation analysis.

Steward et al. (2010) stated that approximately 160 unrelated cases were known to the Barth Syndrome Foundation genetic database, and noted that there were multiple barriers to case ascertainment: the relatively small increase in organic acid excretion is easily missed or may be absent; neutropenia may be intermittent or nonexistent; and a viral etiology for acute CMD is often assumed when CMD is seen in combination with neutropenia, and this misdiagnosis is compounded by the often remarkable improvement in CMD with age, seemingly confirming the suspicion that the patient has recovered from an acute viral insult.

Clinical Management

Ostman-Smith et al. (1994) described a case of type II X-linked 3-methylglutaconic aciduria in a male infant who was admitted to hospital with gross congestive heart failure at the age of 3 weeks. A metabolic cause for his dilated cardiomyopathy was suspected because of the development on the electrocardiogram of an unusual 'camel's hump' shape of the T waves and progressive thickening of the left ventricular wall with increasing echogenicity. Digitalis did not provide sustained improvement and supplementation with L-carnitine was associated with rapid deterioration in cardiac state and may be contraindicated in this condition. At a point when the patient was moribund, large doses of pantothenic acid, a precursor of coenzyme A, produced a dramatic and sustained improvement in myocardial function and in growth, neutrophil cell count, hypocholesterolemia, and hyperuricemia, which suggested that limited availability of coenzyme A was the fundamental pathologic process in this condition. After 13 months, the clinical improvement had been maintained, and myocardial function was nearly normal. Oral pantothenol, unlike pantothenic acid, was not efficacious. Since the specific enzyme defect in this disorder was then unknown, the suggested dietary treatment was entirely empirical.

Nomenclature

Barth et al. (2004) stated that early descriptions of Barth syndrome referred to 'X-linked endocardial fibroelastosis' (EFE) because of the shining pearly aspect of the fibrosis of the endocardium seen at autopsy. However, as methods to visualize the dynamics of the heart in vivo developed, the lack of proper contraction became the focus of attention and the descriptive terminology changed to 'dilated cardiomyopathy.'

Mapping

By means of linkage studies in the large Dutch family reported by Barth et al. (1983), Bolhuis et al. (1991) demonstrated that the BTHS locus is located in Xq28. Multipoint linkage analysis resulted in a maximum lod score of 5.24, with DXS305 being the closest of the markers used. Bolhuis et al. (1991) commented on the large number of genes that have been mapped to Xq28, despite its relatively small physical size, which is estimated to be 5 to 6 Mb.

In a large Australian family in which affected males over 3 generations had dilated cardiomyopathy, short stature, and neutropenia, Ades et al. (1991, 1993) found a maximum lod score of 2.8 at theta = 0.0 with Xq28 polymorphic marker DXS52.

In a large Australian family with X-linked dilated cardiomyopathy, Gedeon et al. (1995) found linkage of the disorder to Xq28, obtaining lod scores of 2.3 at theta = 0.0 with dinucleotide repeat markers near DXS15 and at F8C (300841). The proximal limit of the location of the gene in this family was defined by a recombinant at DXS296, while the distal limit could not be differentiated from the telomere.

In a 4-generation Utah family in which affected males presented with ventricular dysfunction in the first year of life, associated with arrhythmias, heart failure, isolated left ventricular noncompaction, and growth retardation, Bleyl et al. (1997) found linkage to chromosome Xq28, obtaining a maximum lod score of 3.64 (theta = 0) at DXS52. Recombination events narrowed the critical region to an approximately 6.8-Mb interval distal to DSX1193.

Molecular Genetics

In a male proband from each of 4 unrelated families with Barth syndrome, including the large Dutch pedigree originally described by Barth et al. (1981, 1983) and the large Australian family studied by Ades et al. (1993), Bione et al. (1996) identified 4 different truncating mutations in the G4.5 gene (TAZ; 300394.0001-300394.0004). The mutations segregated with disease in each family and were not found in the normal population.

D'Adamo et al. (1997) analyzed the G4.5 gene in 8 additional probands with Barth syndrome and identified mutations in 6 of them (see, e.g., 300394.0006). They also identified a 1-bp deletion (300394.0005) in affected individuals from the large Australian family originally reported by Gedeon et al. (1995) as having X-linked fatal infantile cardiomyopathy, and a missense mutation (300394.0014) in 2 unrelated families diagnosed with endocardial fibroelastosis, 1 of which was the family previously studied by Lindenbaum et al. (1973). D'Adamo et al. (1997) noted that the clinical data on the patients from the latter 3 families was limited and whether other features of Barth syndrome were present could not be established; they suggested that mutations in the G4.5 gene should be considered as a possible cause of infantile CMD affecting males, even in the absence of typical Barth syndrome signs.

In a 4-generation Utah family in which affected males presented with ventricular dysfunction in the first year of life, associated with arrhythmias, heart failure, isolated left ventricular noncompaction, and growth retardation, Bleyl et al. (1997) identified a missense mutation in the G4.5 gene (G197R; 300394.0006) that segregated with disease and was not found in 300 unrelated females. Neutropenia was seen in 2 of the patients, and muscle weakness in 1.

Johnston et al. (1997) evaluated 14 Barth syndrome pedigrees, including the 5 pedigrees previously studied by Kelley et al. (1991) and the 4 families originally reported by Christodoulou et al. (1994), and found mutations in the G4.5 gene in all, including 5 missense mutations (see, e.g., 300394.0006), 4 splice site mutations (see, e.g., 300394.0007), 3 deletions, 1 insertion, and 1 nonsense mutation.

In affected individuals and obligate carriers from 5 unrelated families with Barth syndrome that presented to a hospital in Bristol, England, over a 7-year period, Cantlay et al. (1999) identified mutations in the G4.5 gene (see, e.g., 300394.0006). The authors suggested that Barth syndrome may be more common than previously believed, and concluded that all young male children with idiopathic dilated cardiomyopathy should be investigated for underlying Barth syndrome.

Chen et al. (2002) analyzed the G4.5 gene in 27 Japanese patients with isolated left ventricular noncompaction, including 14 familial cases from 10 families and 13 sporadic cases, and identified a splice site mutation in 1 family (300394.0013) that was not found in 100 Japanese or 100 Caucasian controls. The latter family had a history of unexplained male infant death, with the proband and a distant male relative presenting at 2 months and 3 months of age, respectively, with heart failure. Neither patient nor any other family members had signs of Barth syndrome such as growth retardation or skeletal myopathy. Review of the G4.5 mutations identified to date in 38 reported cases of Barth syndrome and other cardiomyopathies revealed no correlation between location or type of mutation and either cardiac phenotype or disease severity.

Pathogenesis

Schlame and Ren (2006) provided an overview of the molecular basis of Barth syndrome, suggesting that the acyl-specific remodeling of cardiolipin by tafazzin promotes structural uniformity and molecular symmetry among the cardiolipin molecular species, and that inhibition of this pathway leads to changes in mitochondrial architecture and function.

Genotype/Phenotype Correlations

In the families studied by Johnston et al. (1997), no correlation between the location or type of mutation in any of the clinical or laboratory abnormalities of Barth syndrome was found, suggesting that additional factors modify the expression of the Barth phenotype. The clinical histories of most of the subjects investigated by Johnston et al. (1997) had been reported by Kelley et al. (1991) or by Christodoulou et al. (1994). The diagnosis of Barth syndrome was based on the triad of dilated cardiomyopathy, neutropenia, and increased 3-methylglutaconic aciduria in males.

Animal Model

Xu et al. (2006) generated homozygous Drosophila mutants that were unable to express full-length tafazzin and observed an 80% reduction of cardiolipin with diversification of its molecular composition, similar to the changes seen in Barth syndrome patients. Other phospholipids were not affected. Flies with the tafazzin mutation showed reduced locomotor activity, and their indirect flight muscles displayed frequent mitochondrial abnormalities, mostly in the cristae membranes. Xu et al. (2006) concluded that a lack of full-length tafazzin is responsible for cardiolipin deficiency, which is integral to the disease mechanism and leads to mitochondrial myopathy.

Using RNA interference, Acehan et al. (2011) generated tafazzin-knockdown mice, the first mammalian model system for Barth syndrome. Tafazzin-deficient mice developed normally during the first 2 months, but at 8 months they weighed 17% less than control littermates. Tafazzin knockdown resulted in a dramatic decrease of tetralinoleoyl cardiolipin in cardiac and skeletal muscles and accumulation of monolysocardiolipins and cardiolipin molecular species with aberrant acyl groups. Electron microscopy revealed pathologic changes in mitochondria, myofibrils, and mitochondrion-associated membranes in skeletal and cardiac muscles. No overall effect was seen on the measured parameters of cardiac function at 2 months of age in tafazzin-deficient mice, but echocardiography and MRI at 8 months revealed severe cardiac abnormalities, including left ventricular dilation, left ventricular mass reduction, and depression of fractional shortening and ejection fraction.